An AI companion app with 66,000+ contacts was preparing to launch its first paywall. Email would be the primary monetization channel — onboarding users, driving activation, and converting free users into paying customers.
There was just one problem:
The email system was quietly failing underneath the surface.
What I Found
I started with deliverability — because lifecycle strategy means nothing if emails never reach the inbox.
The issues ran deeper than expected:
• 32% of the list was invalid
• 74% of contacts were inactive
• A re-engagement campaign had 174,000 stuck queue entries due to a broken trigger
• Gmail inbox placement had deteriorated significantly
This wasn’t an ESP configuration problem. Authentication was correctly set up. The real issue was operational: poor segmentation, weak sending hygiene, and lifecycle systems running without engagement logic.
The infrastructure existed. The strategy behind it didn’t.
What I Did
Phase 1 — Deliverability Recovery
Conducted a full deliverability audit inside Customer.io, suppressed invalid contacts, rebuilt engagement-based segmentation, fixed the broken re-engagement trigger, and implemented proper sending hygiene across the account.
Result:
Before deliverability
After Deliverability
• +124% Gmail delivery improvement
• -96.3% bounce rate reduction
Phase 2 — Lifecycle System Rebuild
Rebuilt onboarding and retention flows around actual user behavior instead of static time delays.
Users who activated received different journeys than users who stalled. Quiet users were re-engaged early before churn behavior compounded.
The system shifted from mass broadcasting to behavioral lifecycle orchestration.
Onbboarding campaign before and after
Re-engagement campaign before/after
Phase 3 — Paywall Launch Infrastructure
Built the lifecycle communication system supporting the paywall rollout:
• pre-launch education
• launch-day messaging
• post-launch conversion flows
• contingency sequences for non-converting users
The Result
• +124% Gmail delivery improvement
• -96.3% bounce reduction
• 32% invalid contacts identified and suppressed
• 74% inactive audience segmented and addressed
• 174K stuck queue entries fixed
• Full lifecycle infrastructure rebuilt inside Customer.io